Due to the proliferation of Web 2.0, there is a significant increase in the data on the web, which leads to the data overload problem. The potential solution to this issue is Recommendation Systems (RS). RS is available in various domains like music, movies, tourism, etc. The book recommendation domain is less addressed in the literature. Thus, we are proposing a novel hybrid recommendation algorithm that suggests books by employing matrix factorisation techniques. Latent factors are discovered through machine learning algorithms and are embedded to generate the weighted relevance score. On the basis of this relevance score, items are filtered and recommended to the users. The model was evaluated using metrics such as RMSE, Precision@10, Recall@10, and F1 score, and has achieved promising results. Moreover, the novel approach effectively addressed challenges like data sparsity and the curse of dimensionality to a great extent.